[socra]

Hi all,

I'm trying to simulate a Wave with the following properties:

Heigth= 6m
Wavelength: 46m
Period: 5s

The Boundary condition are Velocity inlet at bottom, inlet, lateral and Top. Pressure out at the outlet.

All the boundaries are set with the information of the Fifth order wave.
The mesh is setting Isotropic with a dimension of 0.3m so that I have 20 cells for the height and 150 cells for the wavelength. The timestep is set in order to have CFL<0.5.

The problem is that when I plot the Isosurface of the wave, I have the "non-continuity" of the wave on the boundary. It seems like the wave doesn't slip on the boundary. I attach the screen of the simulation to show the problem.

Anyone can help me?

Thanks!

[Peresvet]

Originally Posted by socra

Hi all,

I'm trying to simulate a Wave with the following properties:

Heigth= 6m
Wavelength: 46m
Period: 5s

The Boundary condition are Velocity inlet at bottom, inlet, lateral and Top. Pressure out at the outlet.

All the boundaries are set with the information of the Fifth order wave.
The mesh is setting Isotropic with a dimension of 0.3m so that I have 20 cells for the height and 150 cells for the wavelength. The timestep is set in order to have CFL<0.5.

The problem is that when I plot the Isosurface of the wave, I have the "non-continuity" of the wave on the boundary. It seems like the wave doesn't slip on the boundary. I attach the screen of the simulation to show the problem.

Anyone can help me?

Thanks!

It should be so. Use side walls. The ratio of wavelength to height is at least 17 times. Second order of sampling in time.

https://youtu.be/e8SWa9MwRJc

[socra]

Originally Posted by Peresvet

It should be so. Use side walls. The ratio of wavelength to height is at least 17 times. Second order of sampling in time.

https://youtu.be/e8SWa9MwRJc

Ok so I need to set wall no-slip to the sides and set 2nd order to discretize the time, right?

[Peresvet]

Originally Posted by socra

Ok so I need to set wall no-slip to the sides and set 2nd order to discretize the time, right?

Yes. You can also do this: left - right - bottom - top = wall.

You can use free surface adaptation by setting free surface volumetric control = 100% and adaptation = 3 or 4 maximum grind level.

[socra]

Originally Posted by Peresvet

Yes. You can also do this: left - right - bottom - top = wall.

Ok, I try with this boundary. Can you explain what do you mean with "The ratio of wavelength to height is at least 17 times." And if you have some papers about it? Thanks a lot!

[Peresvet]

Originally Posted by socra

Ok, I try with this boundary. Can you explain what do you mean with "The ratio of wavelength to height is at least 17 times." And if you have some papers about it? Thanks a lot!

When I used non-automatic grids, there were problems with the amplitude of the wave from time to time. Then I deduced this dependence> 17 times. Length / Height => 17.
It might come in handy.

[socra]

Originally Posted by Peresvet

When I used non-automatic grids, there were problems with the amplitude of the wave from time to time. Then I deduced this dependence> 17 times. Length / Height => 17.
It might come in handy.

So I have a ratio of L/H = 7.67. This could bring problems because the wave is very steep? and because the ratio is < 17 I need an adaptive mesh on the free surface, is that right?

About the adaptive mesh on the free surface, does it double the mesh size level by level? So I have to set the mesh under the last level in order to have the size of the mesh of 0.3m to the free surface?! Or Starccm+ calculate the best size based on my wave? I'm asking because I never used it!

Thanks

[Peresvet]

Originally Posted by socra

So I have a ratio of L/H = 7.67. This could bring problems because the wave is very steep? and because the ratio is < 17 I need an adaptive mesh on the free surface, is that right?

About the adaptive mesh on the free surface, does it double the mesh size level by level? So I have to set the mesh under the last level in order to have the size of the mesh of 0.3m to the free surface?! Or Starccm+ calculate the best size based on my wave? I'm asking because I never used it!

Thanks

Try no adaptation, but there will be more cells.
With the ratio = <17,
the waves will decay along the length of the domain after a short time. Try building on automatic grids.

https://youtu.be/MtjdC5NL2u0

[socra]

Originally Posted by Peresvet

You can use free surface adaptation by setting free surface volumetric control = 100% and adaptation = 3 or 4 maximum grind level.

I set the adaptation. I can set only Volumetric or surface control, what do you mean with "setting free surface volumetric control = 100%"?

[Peresvet]

Originally Posted by socra

I set the adaptation. I can set only Volumetric or surface control, what do you mean with "setting free surface volumetric control = 100%"?

In the mesh generator, free surface volumetric control = 100% of the base value. You will divide by level in the adaptation of grids = physics - model. Volumetric control as a guideline for free surface adaptation in a physics model.

https://youtu.be/2doeE93APO0 --- free surface adaptation

https://youtu.be/H3RXxEL1sG0 --- multi-step mode.

[Peresvet]

Originally Posted by socra

I set the adaptation. I can set only Volumetric or surface control, what do you mean with "setting free surface volumetric control = 100%"?

I do not recommend getting carried away with the adaptation of the free surface in cases of using an overset mesh of moving bodies of complex shapes: ships, airplanes, etc. There is still no perfection, not even close. You can muddy the water with balls, cubes, pipes and nothing more. The resistance characteristics will have a terrible picture, especially the wave resistance characteristics. Forget it. Stir up the water in an empty pool. This is not a temporary recommendation. The overset mesh itself is the source of all design engineer problems.
It is a mistake in cfd to use a movable grid. The resistance in the boundary layer is calculated with high precision by other methods. Error in cfd less than 0.5%? !!! ?? !!! Moreover, all calculations are based on the usual fitting to a practical result. Units on planet Earth are able to count accurately enough without preliminary calculations. For example me. Sorry for being modest.